Black Sigatoka (
Mycosphaerella fijiensis
Morelet) has been causing problems to the national Musacea producers. The development of models, that allow establishing an early prognosis system for the outbreak of the disease, is one of the most effective control measures. The objective of this study was to elaborate a model that would allow in the future the development of a prognostic system for the outbreak of Black Sigatoka in plantains crops in the area. These models relate biological indicators of occurrence of the disease with meteorological data gathered in the Estaciσn Local Chama of Instituto Nacional de Investigaciones Agrνcolas (INIA),. For the development of these models, techniques on data analysis were considered and algorithms that allow the iterative search for correlations between records of consecutive days (temporary windows) of the agro-climatic variables and the biological variables were elaborated. The developed models presented a seasonal dynamics. Different models are obtained if data is considered globally or discriminated according to the dry and wet seasons. Relative humidity, precipitation, wind velocity and solar energy are the variables that best predict the severity of the disease for the dry season 29 days ahead (r
2 = 0.79); while air temperature, evapotranspiration, relative humidity and precipitation are the ones that predict severity 43 days ahead for the rainy season (r
2 = 0.73). Taken globally, it is possible to predict spotted youngest leaf in 23 days ahead using relative humidity and solar energy.